Abstract
Vibrio spp. infections have increased significantly over recent decades. Advances in remote sensing and understanding of environmental parameters driving increased proliferation of pathogenic Vibrio spp. have enabled development of predictive risk models. Here, we implement Bayesian spatial modeling (INLA-SPDE) to forecast infection risks for six vibrio species-V. alginolyticus, V. cholerae non-O1/non-O139, V. fluvialis, V. mimicus, V. parahaemolyticus, and V. vulnificus-and their combined burden, i.e., non-cholera vibriosis, along the United States eastern seaboard. Using reports of vibriosis from 1997 to 2019, we modeled infection risk with a 1-month lead time, assessing 11 environmental and 4 socioeconomic predictors, classifying counties as high, medium, and low risk. The total vibriosis model achieved the highest precision (22.6%) and improved precision over time (slope = 0.71% per year). The predicted probability of vibriosis nearly doubled annually between 1997 and 2019 (93.4% increase). Validation employing 2020-2024 vibriosis data for Florida demonstrated further improvement from the calibration period (slope = 4.37%, precision = 47.0%). Prediction during hurricane season is especially promising, as evidenced for hurricanes Helene and Milton (precision = 65.0%). To assess future climate scenario risk, these models integrated data from Bio-ORACLE v3.0 (CMIP6) under six shared socioeconomic pathways. Results show near-universal, ca. 100%, predicted probability of vibriosis by the end of the century during the peak vibriosis season on the east coast under currently likely scenarios. Given this threat and the promise of the models presented here, development of an early warning system for vibriosis in the eastern United States is deemed critical. IMPORTANCE: This study provides predictive risk models for vibriosis infections along the eastern United States with a 1-month lead time and demonstrates risk of vibriosis under different climate change scenarios. These efforts may lead to the development of future early warning systems, allowing for mitigation of infections and benefit to public health.